The majority of adult HIV infections in sub-Saharan Africa occur in women, highlighting the importance of understanding HIV risk in this population. Numerous studies have suggested that perturbations of vaginal bacterial flora, including bacterial vaginosis (BV), may increase women's risk of HIV acquisition. Recently, molecular techniques for characterization of vaginal flora have enhanced our knowledge of the central role of uncultivated organisms in the microbial ecology of BV. Indeed, several novel bacterial species appear to be substantially more predictive of BV than bacteria previously identified through culture. Understanding the role of vaginal flora in mediating HIV risk is an important HIV prevention goal. The overarching aim of this proposal is to use cutting-edge molecular approaches to test the hypothesis that the presence of specific bacteria in the vaginal flora is associated with increased risk for HIV acquisition. To accomplish this objective, we will conduct a nested case-control study utilizing prospectively collected data from NIH-funded cohorts of women at risk for HIV in Mombasa and Kisumu, Kenya. Vaginal bacterial flora will be evaluated using two distinct and complementary molecular techniques: (1) quantitative 16S ribosomal DNA polymerase chain reaction (PCR) to identify and quantify species that are frequently associated with BV, and (2) broad range PCR with pyrosequencing (deep sequencing) to identify minority species and broad patterns of vaginal bacterial communities in women who acquire HIV versus those who remain HIV-seronegative. The proposed studies will be conducted by an experienced multidisciplinary team including basic scientists, clinicians, epidemiologists, and biostatisticians with an established record of collaborative HIV research in women.